(A CPEP Discussion Paper)
THE VOLCANIC RECORD AND CLIMATIC CYCLES
R. Bryson
PREVIOUS WISCONSIN STUDIES
In the 1970s Dr. Brian Goodman, then a graduate student with a background in
Astronomy, was asked to do a master’s thesis on history of the measured transparency of
the atmosphere (1978). Some partial Soviet studies had been done, such as that of
Pivovarova (1977), and it was known that systematic measurement had begun in 1883 at
Montpelier, in France. It was also known that that the first station for regular
measurement of the atmospheric transparency in the United States had been established
in Madison, Wisconsin about 1904.
This alone seemed to be rationale enough to enlarge the study, but the Center for
Climatic Research had also been engaged in field measurement of atmospheric
transparency from aircraft. This culminated in a round-the-world flight in an
instrumented Navy aircraft and the doctoral dissertation of Dr. Jim Peterson (1968).
In the 1970s, as a CCR project, K. Hirschboeck had scoured the literature for
records of historically observed volcanic eruptions and had put together a list of about
7000 cases (1980). Later this list was almost exactly duplicated by the Smithsonian
Institution (Simkin and Siebert, 1994).
In the 1970s, Bryson and Goodman had been accumulating a database of
radiocarbon dated eruptions, and had developed a database of about 900 cases (1980a,
1980b; Bryson, 1982, 1988) This was used by Goodman to show that most of the
variation of atmospheric transparency could be explained by the volcanic record (1984).
In the 1990s, the base was extended to about 2600 radiocarbon dated eruptions by
Robert Bryson (2006). It is this record that we shall use in this discussion. (See Fig. 1)
It had been generally assumed up to the eruption of Mt. St. Helens in southern
Washington State that the major effect of a volcanic eruption on the transparency of the
atmosphere was due to the “dust” of fine rock particles blasted into the atmosphere by
the volcano (See e.g. Lamb, 1972). This probably led to the concept of only large
eruptions capable of blasting the fine tephra into the stratosphere where a longer fallout
time would produce more climatic effect. After the intense research associated with the
Mt. St.Helens it became clear that it was mostly the very small droplets of saturated
sulfuric acid converted from volcanic (and industrial) sulfur dioxide that had a
sufficiently long half-life in the stratosphere to be climatically important. It was also
established at that time that the sulfuric acid droplets in the “Junge layer” had a very
small absorption to backscatter ratio, and thus always produced cooling.
The sulfur dioxide did not need to be “blasted” into the stratosphere, so the importance
of a volcano became greater as its sulfur dioxide content rose, not by how much solid
tephra was produced.
CURRENT STUDIES
As a consequence, in our investigations here at Wisconsin, we included all sizes
of eruptions, for two medium eruptions might introduce as much or more sulfur dioxide
into the stratosphere as one large eruption. It is also now obvious that the tropopause is
not an ubiquitous barrier, for isentropes from the lower troposphere in lower latitudes
are continuous into the stratosphere in higher latitudes. The author has personally
followed the New York City pollution plume from the ground into the stratosphere over
Iceland.
Fig. 1 The one-century resolution history of the Volcanicity Index, showing the
correspondence of the peaks of global volcanism with the published times of Heinrich
Events and the clear presence of Dansgaard-Oeschger cycles in the volcanic record.
Looking at the Volcanicity Index (VI) record, one is immediately struck with the
apparent regularity of the variations. A spectral analysis or Fourier Transform shows
that the appearance of regularity is indeed borne out. There are very significant
periodicities in the record (Fig. 2). Of particular interest is the extremely significant
concentration of variance in the Heinrich event (HE) frequency range and lesser but still
highly significant variance at the Dansgaard-Oeschger frequency (DO). These are
sufficiently strong that they stand out in figure one without the spectral analysis.
The concentrations of variance at higher frequencies with very little variance
between appear to represent, at least in part, harmonics of the Heinrich and DansgaardOeschger frequencies, suggesting that there is a non-linearity involved, or that it was
introduced in the production of the spectrum. The spectrum was done on the cube root
of the VI to avoid the “ringing” of the spectra done on the raw data. Ringing is what one
gets when doing a Fourier transform on a highly periodic time series that also has a
highly skewed distribution of values.
VARIANCE SPECTRUM OF VOLCANICITY INDEX
VI 99 raw, cube-rooted, 20 century taper
0.04
50 10
VARIANCE
0.03
0.02
14 30
0.01
62 7 55 7
42 7
89 1
20 8
27 1
34 9
25 5
0
0
0.1
0.2
0.3
0.4
0.5
FREQ./CTRY
"White Noise"
5 freq avg
Variance
99 0629
Fig. 2 Fourier transform (spectrum) of the volcanicity time series showing the Heinrich
and Dansgaard-Oeschger signatures, as well as harmonics of the two cycles, beats and
aliases. The whole numbers next to the curve are the periods in years that correspond to
the variance peaks.
It does not seem physically logical that the appearance of the Heinrich and
Dansgaard-Oeschger Events in the volcanic record should be due to oceanic or
atmospheric forcing, but it does make physical sense to say that volcanism, through it
effect on atmospheric transparency might drive those events. This need not be a direct
cause and effect relationship. It is possible through complex oceanic and atmospheric
responses to the modulation of incoming radiation by volcanic aerosols. Some of these
intermediate processes were discussed some time ago by Bond and Lotti (1995).
A SIMPLE APPLICATION OF THE VOLCANIC HISTORY
A simple application of the volcanicity index is a test of its utility in
reconstructing the gross hemispheric mean temperature. This was done by treating the
carbon dioxide content of the atmosphere as a dependent variable and introducing an
atmospheric transparency dependent on the volcanicity index. The translation of the VI
to transparency followed the work of Goodman (1978, 1984). The result is shown in
figure 3.
Fig. 3 Modeled mean surface temperature of the most recent 2.5 millennia with the
transparency of the atmosphere determined by the volcanic aerosol content. The
“present” or zero time on this diagram is 1950 AD.
The results of this simple modeling exercise are very satisfactory in both timing
and magnitude including the recently recognized double nature of the “Little Ice Age”. It
would appear that the Volcanicity Index is a variable to be considered in paleoclimatic
modeling.
REFERENCES
Bond, G.C. and R, Lotti (1995) Iceberg Discharges Into the North Atlantic in Millennial
Time Scales During the Last Glaciation Science 267(5200):1005-1010.
Bryson, R. A., 1982: "Volcans et climat," La Recherche, l3 (135):844-853.
Bryson, R.A., 1988: "Late Quaternary Volcanic Modulation of Milankovitch Climate
Forcing" Theoretical and Applied Climatology 39:115-125.
Bryson, R. A. and B. M. Goodman, 1980a: "Volcanic Activity and ClimaticChanges,"
Science, 27:1041-1044.
Bryson, R.A. and B. M. Goodman, 1980b:"The Climatic Effect of Explosive Volcanic
Activity: Analysis of the Historical Data." Proceedings of a Symposium,
Atmospheric Effects and Potential Climatic Impact of the l980 Eruptions of Mount
St. Helens,, held at Washington, D.C., November 18-19, 1980, NASA Conference
Publication 2240.
Bryson, R.U. and R.A. Bryson (1998). Application of a Global Volcanicity Time-series on
High-Resolution Paleoclimatic Modeling of the Eastern Mediterranean. In “Water,
Environment and Society in Times of Climatic Change,” (A.S. Issar and N. Brown
(eds.), Kluwer Academic Publishers, The Netherlands, pp. 1-19.
Bryson, R. U., R. A. Bryson, and A. Ruter (2006) A calibrated radiocarbon data- base
of late Quaternary volcanic eruptions eEarthDiscussions 1, 123–134
Goodman, Brian M. (1978) Secular variations of the Mean Annual Atmospheric
Transparency in the Northern Hemisphere M.S.Thesis,Department of
Meteorology, University of Wisconsin- Madison, 38pp.
Goodman, B.M. (1984) The Climatic Impact of Volcanic Activity Ph.D.Thesis,Department of
Meteorology, University of Wisconsin-Madison, 245pp.
Hirschboeck, K.K. (1980) A new world-wide chronology of volcanic eruptions.
Paleogeography,
Paleoclimatology, and Paleoecology 2991979/1980):223-241.
Lamb, H.H. (1972) Climate; Present, Past and Future, Vol. 1, London, Methuen & Co
Ltd pp.410-438.
Peterson, J.T. and R.A. Bryson, 1968: "The Influence of Atmospheric Particulates on the
Infrared Radiation Balance of Northwest India", Proc. First National Conf. on
Weather Modification, Albany, N.Y., April 28-May 1, 1968, (Amer. Meteor.
Soc.) pp. 153-162.
Pivovarova, Z.I., 1977: Use of data from surface radiation observations for study of
atmospheric transparency. [English translation of Meteorologiya i Gidrologiya,
9, 328-338.]
Simkin, T. and L. Siebert (1994) Volcanoes of the World Tucson: Geoscience Press
349p.